The German OEM has embraced some sophisticated software systems and combined these with a variety of ‘smart’ robots and virtual reality solutions. The Group uses a cloud-based operating platform, named BMW Services, for central coordination
of autonomous transport systems. The platform simplifies logistics processes significantly. Staff enter driving rules and workflows through an easy-to-use user interface and receive the latest data on all vehicles. Going forward,
the platform will support autonomous transport vehicles built by different manufacturers. The Group is actively involved in the German Association of the Automotive Industry (VDA) and Mechanical Engineering Industry Association
(VDMA), working to establish a standard that will enable communication between all autonomous transport systems in the marketplace. Regardless of the manufacturer, every autonomous transport system should be able to exchange data
with BMW Services.
As well as the AutoTrailer that we featured in last week’s article, a mobile platform which drives underneath a trailer deposited by a truck tractor unit, connects to it and steers it through the plant, the Group has developed several
Following on from the AutoTrailer, another variant of the autonomous outdoor transport system uses devices called AutoBoxes. These autonomous platforms can manoeuvre multiple lattice boxes with components between plant halls. With
a payload of up to 25 tons, they can transport up to 20 lattice boxes at one time. In spring 2019, a pilot application will be launched at the Dingolfing Dynamics Centre. The AutoBoxes will also be introduced at the plants in Shenyang,
China and Berlin next year.
Back in 2015, the BMW Group joined forces with the Fraunhofer Institute IML to develop the first self-driving Smart Transport Robots (STR) for transporting roll containers through logistics areas within production halls. The second
generation is now in operation at BMW Group Plant Regensburg. The flat robots carry roll containers weighing up to one ton and transport them autonomously. They calculate the ideal route independently and move freely through the
plant. The new SLAM (Simultaneous Localisation and Mapping) navigation method does not require permanently installed navigation transmitters in buildings and can therefore be used quickly in a new environment. A built-in battery
module from the BMW i3 powers the STR for a whole work shift.
For delivery of urgent small parts, a smaller version of the STR is used, the so-called miniSTR. This robot transports small load carriers weighing up to 50 kg and calculates its route independently.
The body shop at BMW Group Plant Regensburg has successfully piloted autonomous lift trucks, so-called ‘ants’, over the past twelve months. These autonomous ants bring components from the ‘supermarket’ to the right installation site.
The safety concept includes personal protection and obstacle avoidance, as well as interfaces to other automated industrial trucks. Going forward, a total of eight autonomous ants will supply other areas of series production.
After delivery to the plant, the goods are transported to the assembly line in containers and parts containers of various sizes. For the tiring job of reloading containers from pallets onto conveyor belts or into storage, employees
will be assisted in the future by logistics robots specially developed for this purpose. Four different types of robots, referred to as ‘Bots’ by logistics experts, are currently being tested or have already been integrated into
series production as Marco Prüglmeier, head of innovations and industry 4.0 told me. “When we started on the Connected journey, we quickly learned that one robot would not meet all our requirements so we developed four different
robotic devices. We call these SplitBot, PlaceBot, Pickbot and SortBot.
Stationary ‘SplitBots’ can take full plastic boxes from the pallet in the incoming goods area and place them on a conveyor system that transports the boxes to a warehouse. The SplitBot also makes sure the containers are lined up correctly
for automated storage. Using artificial intelligence, the SplitBot can detect and process up to 450 different containers. Following successful completion of the test operation, plans call for SplitBots to be used in series production
at BMW Group Plant Dingolfing from 2019.
The BMW Group is currently testing the use of a mobile 'PlaceBot' directly on the assembly line. The movable PlaceBot unloads tugger trains and places boxes loaded with goods on a shelf. It uses an image recognition system to classify
these small load carriers and determine the ideal grip point from the combined input of sensor technology, camera and artificial intelligence. It can also move autonomously in a predetermined area.
Another logistics robot, the‘PickBot’, developed with machinery supplier FESTO, collects various small parts from appropriate supply racks. Different parts are recognised by a self-developed, self-trained artificial intelligence, using
a camera to scan items to make up ‘car set’ assembly kits. The PickBot then calculates the right grip point and uses suction cups and robotic grippers to pick up the items. Over the long term, the PickBot should be able to recognise
and handle up to 50,000 small parts and will likely be used in series production at BMW Group Plant Leipzig from 2019 on.
The 'SortBot' is already deployed in series production at Plant Leipzig stacks empty containers on pallets before they re-enter circulation. These Three SortBots will be in use at BMW Group Plant Leipzig by the end of the year.
Gloves with integrated scanners and displays, data glasses and smart watches are increasingly used by the Group, to support logistics employees. The transition to paperless logistics, with digitally labelled containers and shelves,
opens up new areas of application for mobile devices. Glove scanners read the electronic label and indicate the exact contents of the small load carrier on a small display that can be worn on the arm.
One aim of paperless logistics is for all containers to have a single label. This ‘unique container ID’ will contain just a single QR code and a number. All information relating to the supplier, content or storage location is stored
centrally and can be called up by various scanners. A special logistics app shows the employee all relevant information for their tasks on a smartphone – for example, the position of a small load carrier and whether it has the
The use of augmented reality glasses helps the employee sort parts into the right order. They see which component goes into which shelf in the data glasses’ field of vision. Correct work steps are confirmed in green, while errors are
visually highlighted. The use of data glasses therefore helps staff keep track of the large number of different components and avoid picking mistakes.
The use of virtual reality already plays an important role in planning logistics spaces. In a virtual environment, planners can quickly and efficiently lay out future logistics areas completely and assess how much space is needed,
for instance. Planning is based on 3D data representing the real structures of a logistics hall. For the past several years, the BMW Group has been capturing its plants in digital form with millimetre accuracy, using special 3D
scanners and high-resolution cameras. This creates a three-dimensional image of the structures, so that manual recording on site is no longer needed. When planning future logistics areas, BMW Group experts can now combine existing
data with a virtual 'library' of shelves, lattice boxes, small load carriers and around 50 other widely-used operating resources.
The selection, placement, movement and removal of logistics structures and areas, as well as distance and area measurements, can be simplified in the virtual environment. Several planners can also work on the same area design at the
same time, regardless of their location.
To ensure the data used is exactly right, the BMW Group is also relying on artificial intelligence. Artificial, neural networks are first 'trained' by presenting the computer with tasks and results to learn from – for example, the
system familiarises itself with the appearance of different container types from photos taken from different angles. By training the network in this way, it becomes capable of recognising the container types it has learned from
new photos reliably and independently. Applications incorporating neural networks require enormous computing power. To secure this, the BMW Group is investing in state-of-the-art computer hardware and software.
Storage areas and their aisles occupy a great deal of space at any plant. In conjunction with two external partners, BMW Group Plant Regensburg is currently testing a modern small load carrier that also comes with a pneumatic gripper
robot. What is unique about this approach is the idea of simply stacking same-sized containers in a grid construction above and next to each other, thereby reducing the amount of space needed. Small transport robots drive on top
of the grid and remove the goods needed by rearranging the containers. As a result, containers with frequently requested goods automatically land higher up and other containers lower down.
A pneumatically-powered lightweight robot then removes the parts from the containers. Using a special gripper, it also takes sensitive parts out of boxes and gets them ready for use as a pre-configured assembly package. Thanks to a
well thought out safety system, the robot is able to work alongside employees without a protective barrier.